The Environmental Impacts on the Star Formation Main
Sequence Tackled by MAHALO-Subaru Campaign
Yusei Koyama (NAOJ)
Subaru user’s meeting (2014/1/19-21)
T. Kodama (NAOJ), M. Hayashi (ICRR/Tokyo), K. I. Tadaki (NAOJ),I. Tanaka (NAOJ), R. Shimakawa (Sokendai), M. Tanaka (NAOJ),
I. Smail (Durham), J. Kurk (MPE), D. Sobral (Leiden/Lisbon), M. Swinbank (Durham), J. Geach (Hertfordshire), P. Best
(Edinburgh)
MAHALO-Subaru collaboration
Star Formation “Main Sequence”
Elbaz et al. (2007)
= SFR-M★ relation for star-forming galaxies
Main Sequence (normal
SF)
Starburst
Passive
log M ★ [M]
SF
R [
M/y
r]
From NEWFIRM medium-band survey (Whitaker et al. 2012)
SF main sequence out to z>2
Galaxy properties vs. Environment
ES0
Spiral
High-density
“ Morphology–Density Relation”
red, old, low SF activity
blue, young, high SF activity
Low-density
(Dressler 1980)
Little environmental impacts ?
Local star-forming galaxies from SDSS (Peng et al. 2010)
low-density(z=0)
log M★
SF main sequence is “independent” of environment at z=0
high-density(z=0)
log
SF
R [
M
/yr]
low densit
y
Field Cluster
high densit
y
Little environmental impacts ?
Q: How about in distant universe?
MAHALO-Subaru project
Narrow-band Ha/[OII] emission-line survey for 0.4<z<2.5
MApping H-Alpha and Lines of Oxygen with Subaru
Collaborators: T.Kodama (PI), M.Hayashi, K.Tadaki, I.Tanaka, R.Shimakawa
Tadaki+’12Koyama+’13a
Hayashi+’12
Tadaki+’13,14
,14Koyama+ in prep
Dec. 2013
High-z structures revealed by MAHALO
(Koyama+11)
(Koyama+10)
(Hayashi+10, 11)
(Tadaki+12)
(Koyama+13)
(Hayashi+12)
z=0.4
z=0.8
z=1.5
z=1.6
z=2.2
z=2.5
Many strong emitters
in z>1.5 cluster cores
HiZELS: High-Z Emission-Line Survey
Collaborators: I. Smail, P. Best, D. Sobral, J. Geach, M. Swinbank
H a@
z=0.4
H a@
z=0.8
H a@
z=1.5
H a@
z=2.2Total ~2 deg2 survey in COSMOS & UDS
Sobral et al. (2013)Subar
u filter
UKIRT filter
~500-2000 Ha emitters at each redshift,providing excellent comparison sample for our MAHALO cluster samples.
Cluster vs. Field comparison out to z~2
Important Notes:
1) We select galaxies with EWr >30A as SF galaxies.
2) We use galaxies Rc < 2 Mpc (physical) as “cluster” galaxies.
z T(look-back)
Cluster(MAHALO
)
Field(HiZELS)
log M(lim)
SFR(lim)
0.4 4 Gyr CL 0939(Ngal=80)
NBz(Ngal=252
)9.0 0.3
0.8 7Gyr RXJ1716(Ngal=34)
NBJ
(Ngal=424)
9.8 4.0
2.2 10 Gyr PKS1138(Ngal=48)
NBK
(Ngal=583)
10.0 15.0
(Koyama et al. 2013b, MNRAS, 434, 423)
Cluster vs. Field comparison out to z~2
(Koyama et al. 2013b, MNRAS, 434, 423)
Ha emitters with EWr > 30A
Main sequence is always independent of environment since z~2 !
Interpretation: rapid SF quenching
Cluster
Interpretation: rapid SF quenching
rapid quenching
slow quenching
M★
SF
R
M★
SF
R
: normal SF galaxy: transitional galaxy: passive galaxy
✗✓
M★, SFR, ΔMS distribution
: Field
: Cluster(MAHALO)
(HiZELS)
Excess of massive
galaxies in proto-cluster PKS1138-262
(Koyama et al. 2013b)M
SF
R
M
SF
RM
SF
R
Our MOIRCS+NB(Ha) survey revealed red Ha emitters occupy the core of z=2.16 proto-cluster (PKS1138-262).
Massive SF galaxies in z>2 proto-cluster
(Koyama et al. 2013a)
: 24um-source
Red emitters are massive (M ★ >1011M), and clearly dominate dense environment at z~2.
■: red HAE (J-K>1.4)■: green HAE (0.8<J-K<1.4)■: blue HAE (J-K<0.8)
Dust extinction vs. environment
(Koyama et al. 2013b)
Dust extinction may be higher in high-density environments.
z=0.4 sampleA(Ha) from
SFR(IR)/SFR(Ha)
A(Ha) fromA(Ha)-M★ relation
(Garn & Best 2010)
(Garn & Best 2010)
AHa-M★ relation (z=0)
Conclusions (1) With MAHALO-Subaru campaign, we find that SF main sequence is independent of environment at any time in the history of universe since z~2, suggesting rapid SF quenching.
(2) SF galaxies in z=2.2 proto-cluster environments tend to be more massive (and showing redder J-K colours) than the field counterparts at the same redshift.
(3) SF galaxies “surviving” in high-density environments tend to be more highly obscured by dust, suggesting a different mode of SF activity in clusters/groups.
M★ distribution “along” the MS does depend on environment.
Environment may affect dust properties (SF geometry or mode).
Cluster vs. field difference is always small (0.2 dex at maximum)